
% Author: Kazuma Takakura
% Date: 17 May 2025
% Time: 10:51:33

\begin{table}[h!]\footnotesize
  \centering
  \caption{Summary Statistics (Matched Sample)}
\label{tab:sumstat_psm}
\scalebox{1}{
\begin{threeparttable}
\begin{tabular}{lccccc}\toprule
 Dependent Variable & Treatment &  Control  & Difference & N   \\\midrule\midrule
 Panel A: Baseline & & & &   \\ 
 DT score^{a} & -0.103 & 0.048 & -0.150 & 192  \\ 
                               & [ 1.097 ] & [ 0.967 ] & ( 0.558 ) &   \\ 
                               &   &   & \{ 0.893 \} &   \\ 
 RSES score^{a} & -0.003 & -0.057 & 0.054 & 196 \\ 
                               & [ 1.042 ] & [ 1.039 ] & ( 0.772 ) &   \\ 
                               &   &   & \{ 0.948 \} &   \\ 
 CPCS score^{a} & -0.008 & -0.199 & 0.191 & 196 \\ 
                               & [ 0.979 ] & [ 1.073 ] & ( 0.382 ) &   \\ 
                               &   &   & \{ 0.806 \} &   \\ 
 Household size & 4.357 & 4.265 & 0.092 & 196  \\ 
                               & [ 1.229 ] & [ 1.198 ] & ( 0.634 ) &   \\ 
                               &   &   & \{ 0.948 \} &   \\ 
 Household head age & 47.306 & 46.689 & 0.617 & 196 \\ 
                               & [ 8.804 ] & [ 9.409 ] & ( 0.690 ) &   \\ 
                               &   &   & \{ 0.948 \} &   \\ 
 Household head education & 2.663 & 3.163 & -0.500 & 196  \\ 
                               & [ 3.315 ] & [ 3.531 ] & ( 0.248 ) &   \\ 
                               &   &   & \{ 0.893 \} &   \\ 
 \\ 
 Panel B: Follow-up & & & &   \\ 
 School attendance & 0.571 & 0.531 & 0.041 & 196  \\ 
                               & [ 0.497 ] & [ 0.502 ] & ( 0.684 ) &   \\ 
                               &   &   & \{ 0.818 \} &   \\ 
 Grade repeat & 0.051 & 0.041 & 0.010 & 196 \\ 
                               & [ 0.221 ] & [ 0.199 ] & ( 0.670 ) &   \\ 
                               &   &   & \{ 0.818 \} &   \\ 
 Drop out & 0.357 & 0.449 & -0.092 & 196  \\ 
                               & [ 0.482 ] & [ 0.500 ] & ( 0.344 ) &   \\ 
                               &   &   & \{ 0.637 \} &   \\ 
 Tutoring & 0.367 & 0.459 & -0.092 & 196  \\ 
                               & [ 0.485 ] & [ 0.501 ] & ( 0.278 ) &   \\ 
                               &   &   & \{ 0.637 \} &   \\ 
 Self-study & 0.469 & 0.429 & 0.041 & 196 \\ 
                               & [ 0.502 ] & [ 0.497 ] & ( 0.722 ) &   \\ 
                               &   &   & \{ 0.811 \} &   \\ 
 Rapid math test score^{a} & -0.252 & 0.136 & -0.389**  & 196  \\ 
                               & [ 1.094 ] & [ 0.873 ] & ( 0.034 ) &   \\ 
                               &   &   & \{ 0.203 \} &   \\ 
 RSES score^{a} & 0.349 & -0.232 & 0.581** & 190  \\ 
                               & [ 1.029 ] & [ 0.923 ] & ( 0.024 ) &   \\ 
                               &   &   & \{ 0.002 \} &   \\ 
 CPCS score^{a} & 0.344 & -0.255 & 0.598**&  190 \\ 
                               & [ 1.020 ] & [ 0.936 ] & ( 0.014 ) &   \\ 
                               &   &   & \{ 0.001 \} &   \\ 
\midrule
\end{tabular}
\begin{tablenotes}
\item (a) Standardized in the follow-up sample. 
\item (b) For the propensity score matching, we use covariates including student's grade, sex, branch dummy (location), parents' income source, last income per family member, number of household members, age of household head, education level of household head, teacher's age, sex, and phone survey dummy.
\item (c) Standard deviations are reported in square brackets. Standard errors, which are reported within parentheses are clustered by school at the baseline. 
\item (d) Romano-Wolf multiple hypothesis testing p-values are reported in curly brackets.
Statistical significance is indicated by stars based on the wild clustered bootstrap p-values reported in parentheses: $*$ denotes significance at the 10\% level, $∗∗$ at the 5\% level, and $∗∗∗$ at the 1\% level. 
\end{tablenotes}
\end{threeparttable}
}
\label{tab:addlabel}%
\end{table}




